メニュー
When a task can be parallelized (bulk code review, multi-module refactor, batch analysis) → split into subtasks running in isolated git worktrees via multiple workers.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
When prompt, tool, or agent logic changes need regression testing → run eval benchmarks to detect regressions or improvements.
When the user wants to research academic papers → search ArXiv, extract PDFs, summarize/review/Q&A, generate experiment code.
When you need to interact with web pages using the user's logged-in browser → control Chrome via @playwright/cli daemon, reusing cookies/session with persistent tabs.
When user needs AnyGenIO content generation → unified CLI for task execution with progressive disclosure.
When you need to persist files (reports, docs, evidence) beyond the session → create/query/delete durable artifacts.
When the user wants text-to-speech audio → generate voice files locally (macOS say + afconvert, outputs m4a).
| name | openmax |
| description | When a task can be parallelized (bulk code review, multi-module refactor, batch analysis) → split into subtasks running in isolated git worktrees via multiple workers. |
| triggers | {"intent_patterns":["openmax|parallel review|并行|多worker|多任务|code review|批量","同时.*处理|handle.*simultaneously|一起.*跑|run.*parallel","批量.*审查|bulk.*review|全部.*模块|all.*modules","拆分.*任务|split.*task|分.*子任务|break.*subtasks","多个.*文件.*同时|multiple.*files.*parallel|加速.*review","worktree|隔离.*执行|isolated.*execution"],"context_signals":{"keywords":["openmax","parallel","review","并行","worktree","批量","同时","拆分","隔离","bulk","concurrent"]},"confidence_threshold":0.7} |
| priority | 9 |
| requires_tools | ["bash"] |
| max_tokens | 300 |
| cooldown | 30 |
| capabilities | ["code_edit","code_review","research","analysis","parallel"] |
| activation_mode | explicit |
| output | {"format":"markdown","artifacts":true,"artifact_type":"document"} |
在隔离 git worktree 中启动 N 个 CC worker，并行执行任务，汇总 .openmax/reports/ 下的产出。
claude CLI 在 PATH 中.openmax/briefs/<task>.md# 1) 基于已有 briefs 并行启动所有 worker
python3 skills/openmax/run.py dispatch --tasks "research-codebase,review-perf,review-arch"
# 2) 指定 goal，自动生成 briefs 并启动
python3 skills/openmax/run.py dispatch --tasks "review-lark,review-eval" --goal "Review the elephant.ai codebase"
# 3) 查看 worker 状态（检查进程 + 报告）
python3 skills/openmax/run.py status
# 4) 汇总已完成的报告
python3 skills/openmax/run.py collect
| 参数 | 说明 |
|---|---|
--tasks | 逗号分隔的任务名列表（对应 .openmax/briefs/<name>.md） |
--goal | 可选：补充到每个 brief 的总目标说明 |
--brief-dir | brief 所在目录，默认 .openmax/briefs |
--report-dir | 报告输出目录，默认 .openmax/reports |
--worktree-base | worktree 根目录，默认 .openmax-worktrees |
--dry-run | 只输出计划，不执行 |
.openmax/briefs/<task>.mdgit worktree add -b openmax/<task> .openmax-worktrees/openmax_<task> mainclaude --dangerously-skip-permissions --print <brief>collect 汇总 .openmax/reports/*.md